eyelinesOnline: Eyelines Online

###################################################
# path - full path to folder containing .edf's, .r2e's and meta.txt (ending with "/")
# fixation_threshold - actual fixation threshold in training games
# left_border, epoch_size - left and right borders of classifier epoch
# channels - numbers of channels used in experiment
# A1, A2 - idex numbers of reference electrodes !in channels list!
# low, high - filters
###################################################
buildClassifier <- function(path, epoch_size=1000, 
                            left_border=-500, channels = c(1:5,7,9:21), A1=22, A2=23, 
                            low=F, high=30, bsln_start = 200, bsln_end = 300,
                            sRate = 500, times_seq = seq(300,450, 20), decimation_window = 50,
                            no_button_press = F, random_non_target = F, ball_only = F,
                            n_random_nontarget = F)
  {
  #channels from Trofimov's clf = c('PZ','P3','P4','P1','P2','PO7','PO8','PO3','PO4','Oz','O1','O2','POz')
  
  # !!!TMP!!!
#   f_channels <- c(7,8,9,10,11,12,13,14,15,16,17,18,19);
  # !!!TMP!!!
  

  l <- load_eye_data(paste0(normalizePath(path), '/'), epoch_size, left_border, sRate,
                    channels, A1,A2, low, high, no_button_press, random_non_target, ball_only, n_random_nontarget)
  
  ret <- process_params(l, channels, A1,A2, low, high, bsln_start, bsln_end, left_border, times_seq, decimation_window)
  
  class(ret) <- 'eyelinesOnline_eeg_classifier'
  
  ret
}


print.eyelinesOnline_eeg_classifier <- function(cls){
  
  con <- file()
  res <- cls
  dump('res', file=con)
  classStr <- paste(readLines(con), sep = '', collapse='\n')
  
  on.exit(close(con))
  
  cat(sprintf(
    '
#################################################################
# 
# Target epochs:    %i
# Nontarget epochs: %i
#
# Specifity:
#   Train: %.3f ± %.3f
#   Test:  %.3f ± %.3f
# Sensivity:
#   Train: %.3f ± %.3f
#   Test:  %.3f ± %.3f
# Accuracy:
#   Train: %.3f ± %.3f
#   Test:  %.3f ± %.3f
# AUC:
#   Train: %.3f ± %.3f
#   Test:  %.3f ± %.3f
#   All:   %.3f ± %.3f
#################################################################

library(Resonance)
library(Resonate)
library(eyelinesOnline)
    
%s

online_epoch_start <- min(res$bsln_start, res$times_seq[1])
online_epoch_end <- max(res$bsln_end, (res$times_seq + res$decimation_window))
    
refs <- c(res$A1, res$A2)

online_epoch_size <- online_epoch_end - online_epoch_start
online_epoch_shift <- online_epoch_start - 500

times_seq <- res$times_seq - online_epoch_start

bsln_start <- res$bsln_start - online_epoch_start
bsln_end <- res$bsln_end - online_epoch_start

process = function(){

FS <- signalPreparation(input(1), low=res$low, high=res$high, notch=50, refs=refs, channels=res$channels)
ev <- perform_time_correction(input(1), input(2))
RA2 <- cross.windowizeByEvents(FS, ev, online_epoch_size/1000*SI(FS)$samplingRate, shift=online_epoch_shift/1000*SI(FS)$samplingRate)
RA3 <- pipe.medianWindow(RA2, (bsln_start)/1000* SI(RA2)$samplingRate, (bsln_end)/1000* SI(RA2)$samplingRate)
RA4 <- pipe.trof.classifier2(RA3, res$W, res$th, times_seq/1000, res$decimation_window/1000)
RA5 <- filter_fast_events(RA4)
createOutput(RA5,"RES")
}',
    res$target_epochs, res$nontarget_epochs,
    
    cls$quality$spec$tr[1],  cls$quality$spec$tr[2],
    cls$quality$spec$tst[1], cls$quality$spec$tst[2],
    
    cls$quality$sens$tr[1],  cls$quality$sens$tr[2],
    cls$quality$sens$tst[1], cls$quality$sens$tst[2],
    
    cls$quality$acc$tr[1],  cls$quality$acc$tr[2],
    cls$quality$acc$tst[1], cls$quality$acc$tst[2],
    
    cls$quality$auc$tr[1],  cls$quality$auc$tr[2],
    cls$quality$auc$tst[1], cls$quality$auc$tst[2],
    cls$quality$auc$all[1], cls$quality$auc$all[2],
    
    classStr
  ))
  
  invisible(cls)
}